Recognition of nonpalindromic DNA sequence by a peptide

Yasunori Aizawa, Yukio Sugiura, Masaru Ueno, Yasuo Mori, Keiji Imoto, Keisuke Makino, and Takashi Morii. Biochemistry 1999 38 (13), 4008-4017...
0 downloads 0 Views 855KB Size
11137

J. Am. Chem. SOC. 1994,116, 11137-11138

Communications to the Editor Recognition of Nonpalindromic DNA Sequence by a Peptide Heterodimer with Artificial Dimerization Module

C23Ad G23Ad G23Ad C23Ad G23Cd C23Cd G23Cd C23Cd

----

1 2 3 4 5 6 7 8 9101112131415

Masaru Ueno, Mamoru Sawada, Keisuke Makino, and Takashi Morii* Department of Polymer Science & Engineering Kyoto Institute of Technology Matsugasaki, Sakyo-ku, Kyoto 606, Japan Received July 6, 1994 Sequence-specific DNA binding of the gene regulatory protein is often mediated by dimeric species. Recent studies on the action of basic-helix-loop-helix and basic-leucine-zipper (bZIP) proteins have demonstrated an important role of the heterodimers in the conirol of specific gene activation.’ While the mechanisms by which protein homodimers recognize the palindromic DNA sequences have emanated from the X-ray crystallographic studies of many protein-DNA complexe~,*~~ it has yet to be established whether the protein heterodimers could recognize nonpalindromic DNA sequences4 The heterodimer formation is a consequence of the specific protein-protein interaction between two different monomers. Thus, a prerequisite for DNA binding of the heterodimer is a module that appropriately controls such specific protein-protein interactions. We report here a sequence-specific DNA binding of peptide heterodimers with an artificial dimerization module5 consisting of a P-cyclodextrin (P-Cd) and its guest compound. Because it is the host-guest inclusion complex that regulates the dimerization, attaching P-Cd to one peptide and a guest molecule to another peptide will allow a specific formation of peptide heterodimer. We used peptides corresponding to the basic region of the yeast transcriptional activator GCN46 (G23) and an enhancer binding protein CEBP7 (C23) to study the DNA binding of heterodimer, since the basic region peptides of bZIP proteins alone are sufficient for the sequence-specific DNA binding when covalently*-’ or noncovalently dimerized.s Modification of the C-terminal cysteines of the peptides with mono-6-deoxy6-iodo-~-cyclodextrin(Cd) or N-bromoacetyl- 1-adamantanemethylamine (Ad) afforded four different peptides (G23Ad, G23Cd, C23Ad, and C23Cd) that are capable of forming specific homo- (G23Ad/G23Cd and C23Ad/C23Cd) and heterodimers (C23Ad/G23Cd and C23Cd/G23Ad).5.12 GCN4 and C E B P are *-To whom- corresponderxe should be addressed. Present address: Institute for Chemical Research, Kyoto University. TJji,Kyoto 61 1, Japan. (1) Jones, N. Cell 1990, 61, 9 and references therein. (2) Harrison, S. C.; Aggarwal, A. K. Annu. Rev. Biochem. 1990, 59, 933. (3)Pabo, C. 0.; Sauer, R. T. Annu. Rev. Biochem. 1992, 61, 1053. (4) Vinson, C. R.; Hai, T.; Boyd, S . M. Genes Dev. 1993, 7, 1047. (5) Ueno, M.; Murakami, A.; Makino, K.; Morii, T. J . Am. Chem. SOC. 1993, 115, 12575. (6)Hinnebush, A. G. Proc. Natl. Acad. Sci. U S A . 1984, 81, 6442. (7) Landschultz, W. H.; Johnson, P. F.; Adashi, E. Y.; Graves, B. J.; McKnight, S . L. Genes Dev. 1988, 2, 786. (8) (a) Talanian, R. V.; McKnight, C. J.; Kim, P. S . Science 1990, 249, 769. (b) Talanian, R. V.; McKnight, C. J.; Rutkowski, R.; Kim, P. S . Biochemistry 1992, 31, 687 1. (9) (a) Cuenoud, B.; Schepartz, A. Science 1993,259,510. (b) Cuenoud, B.; Schepartz, A. Proc. Natl. Acad. Sci. U S A . 1993, 90, 1154. (10) (a) Park, C.; Campbell, J. L.; Goddard, W. A., 111. Proc. Natl. Acad. Sci. USA. 1992, 89, 9094; (b) 1993, 90, 4892. (1 1) Morii, T.; Shimomura, M.; Morimoto, M.; Saito, I. J . Am. Chem. SOC. 1993, 115, 1150. (12)The amino acid sequences of C23 and G23 are Ac-NEYRVRRERT;NIAVRKSRDKAKQC-NH2 and Ac-DPAALKRARNTEAARRSRARKLQC-NH2, respectively. Peptides G23 and C23 were synthesized as described previously (refs 5 , 11). Structures of C23, C23Ad, and

0002-7863/94/1516-11137$04.50/0

3’

T A A C A G T A 5’

Figure 1. DNase I f ~ o t p r i n t i n g pattern ’~ of the nonpalindromic CE/ CR sequence by the homo- and heterodimers, indicating specific binding of the heterodimers C23AdG23Cd and C23Cd/G23Ad at the CE/CR sequence. The bracket denotes the nonpalindromic CE/CR sequence. Lane 1, no peptide; lane 2, Maxam-Gilbert chemical G A reaction; lanes 3, DNase I digestion with no peptide; lanes 4-6, C23Ad/G23Cd (50, 150, and 300 nM); lanes 7-9, C23Cd/G23Ad (50, 150, and 300 nM); lanes 10-12, G23Ad/G23Cd (50, 150, and 300 nM); and lanes 13-15, C23Ad/C23Cd (50, 150, and 300 nM). DNA was 5’-endlabeled with polynucleotide kinase.I5 Reaction mixture contained 20 mM Tris-HC1 (pH 7 . 3 , 4 mM KC1, 2 mM MgC12, 1 mM CaC12, 20 p M poly(d1-dC), 20 000 cpm singly S-”P-end-labeled 46-mer DNA (5'-TCG AATTCCACAGTGAGAAATGACGC AATCCAGACTAGGATCCGC-3’, double-stranded) containing CE/CR sequence, and 1:1 peptide dimer where indicated, in 20 pL total volume. Nuclease digestion was initiated on addition of 1 unit of DNase I on ice and was quenched by addition of 20 ,uL of 0.6 M NaOAc, 10 mM EDTA, and 0.4 mM calf thymus DNA after 30 s. Samples were purified by phenol-chloroform extraction and ethanol precipitation, suspended in 80% formamide loading dye, run on a 12% sequencing acrylamide gel, and analyzed by autoradiography.

+

known to recognize palindromic sequences with a half-site of 5’-ATGAC-3’ (CRE) and 5’-ATTGC-3’ (CE), re~pectively.’~ Combination of these half-sites gives a nonpalindromic sequence (CEKR), 5’-ATGACGCAAT-3’, for the target of the peptide heterodimer. Specific recognition of the nonpalindromic DNA sequence by heterodimers was tested by using deoxyribonuclease (DNase) I f~otprinting’~ (Figure 1). Both C23Ad/G23Cd (lanes 4-6) and C23CdG23Ad (lanes 7-9) show specific protection at the CEKR sequence. In contrast, G23AdG23Cd partially protects the canonical ATGAC half-site of the CEKR sequence (lanes 11 and 12), and no protection is observed at the 3’-side of the C23Cd were characterized byH NMR spectroscopy and amino acid analysis. Peptide concentrations were determined by quantitative amino acid analysis with a-aminobutyric acid as an internal standard. Abbreviations for the amino acids are A, Ala; C, Cys; D, Asp; E, Glu; I, Ile; K, Lys; L, Leu; N, Asn; P, Pro; Q, Gln; R, Arg; S , Ser; T, Thr; V, Val; Y, Tyr. (13) Kerppola, T. K.; Curran, T. Curr. Opin. Struct. Biol. 1991, 1, 71. (14) Gallas, D. J.; Schmitz, A. Nucleic Acids Res. 1978, 5, 3157. (15) Maniatis, T.; Fritsch, E. F.; Sambrook, J. Molecular Cloning: A Laboratory Manual; Cold Spring Harbor Laboratory: Cold Spring Harbor, NY, 1987.

01994 American Chemical Society

11138 J. Am. Chem. SOC.,Vol. 116, No. 24, 1994

E

Communications to the Editor

0.4

.-f 0.2 t 0.0

;:::I:;li .-

~

,&,;

0 ,, o

T, 0.2

;0.0

Io 9

10.8

10"

Peptide (M)

Figure 2. Titration of gel shift" for the basic region heterodimer C23CdG23Ad (A) and the homodimer C23AdC23Cd (B) to CECR21 (O), CRE2l (0),and CE21 (A) oligonucleotides. Extent of saturation for double-stranded oligonucleotides CECR21 (5'-CGGATTGCGTCAT)TT"C-3'): CRE21 (5'-CGGATGACGTCATTTTTTTTC-3'), and CE21 ( 5 ' - C G G A T T G C G C A A T C - 3 ' ) was obtained by averaging data from duplicate experiments, Binding reactions were-carried-out in the presence of the indicated peptides with